Automated Self-Administered 24-H Dietary Assessment Tool (ASA24) recalls for parent proxy-reporting of children's intake ( 4 years of age): a ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 https://doi.org/10.1186/s40814-021-00864-6 RESEARCH Open Access Automated Self-Administered 24-H Dietary Assessment Tool (ASA24) recalls for parent proxy-reporting of children’s intake (> 4 years of age): a feasibility study Isobel Sharpe1, Sharon I. Kirkpatrick2, Brendan T. Smith3,4, Charles D. G. Keown-Stoneman5,6, Jessica Omand7, Shelley Vanderhout5,8, Jonathon L. Maguire5,8,9, Catherine S. Birken7,8, Laura N. Anderson1,7* and on behalf of the TARGet Kids! collaboration Abstract Background: Robust measurement of dietary intake in population studies of children is critical to better understand the diet–health nexus. It is unknown whether parent proxy-report of children’s dietary intake through online 24-h recalls is feasible in large cohort studies. Objectives: The primary objective of this study was to describe the feasibility of the Automated Self-Administered 24-h Dietary Assessment Tool (ASA24) to measure parent proxy-reported child dietary intake. A secondary objective was to compare intake estimates with those from national surveillance. Methods: Parents of children aged 4–15 years participating in the TARGet Kids! research network in Toronto, Canada were invited by email to complete an online ASA24-Canada-2016 recall for their child, with a subsample prompted to complete a second recall about 2 weeks later. Descriptive statistics were reported for ASA24 completion characteristics and intake of several nutrients. Comparisons were made to the 2015 Canadian Community Health Survey (CCHS) 24-h recall data. Results: A total of 163 parents completed the first recall, and 46 completed the second, reflecting response rates of 35% and 59%, respectively. Seven (4%) first recalls and one (2%) second recall were excluded for ineligibility, missing data, or inadvertent parental self-report. The median number of foods reported on the first recall was 18.0 (interquartile range (IQR) 6.0) and median time to complete was 29.5 min (IQR 17.0). Nutrient intakes for energy, total fat, protein, carbohydrates, fiber, sodium, total sugars, and added sugars were similar across the two recalls and the CCHS. Conclusions: The ASA24 was found to be feasible for parent proxy-reporting of children’s intake and to yield intake estimates comparable to those from national surveillance, but strategies are needed to increase response rate and support completion to enhance generalizability. Keywords: Measurement, Child, Parents, Self-report, Nutrition assessment, Nutrition surveys * Correspondence: LN.Anderson@mcmaster.ca 1 Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, Ontario, Canada 7 Division of Child Health Evaluative Sciences (CHES), Sick Kids Research Institute, Toronto, Ontario, Canada Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 2 of 10 Key messages regarding feasibility dietary intake is necessary for informing guidelines and in- terventions to improve health. What uncertainties existed regarding the feasibility? Self-reported measures are commonly used to quantify The Automated Self-Administered 24-H Dietary As- dietary intake in nutritional epidemiology [7]. These sessment Tool (ASA24) is a self-reported 24-h recall measures are prone to measurement error from a developed by the US National Cancer Institute. range of sources, including imperfect memory [8]. With the convenience of an online system, it has However, as compared with objective measures, which broad applications for collecting nutritional informa- are relatively rare in the field of dietary assessment as tion within epidemiological cohorts; however, there well as burdensome and expensive, self-reported tools are limited reports of its feasibility as a tool for par- provide the ability to collect data on a wide range of ent report of child intake. Previous studies have foods and drinks and offer relative ease of administra- found the ASA24 to be feasible among adults and tion [7, 9]. Food frequency questionnaires (FFQs) adolescents; however, little research has focused on have often been used in epidemiologic studies due to younger children. The feasibility considerations for their low cost, but validation research suggests that young children, especially those under 10 years of 24-h recalls capture intake with less systematic error age, are unique given that parent proxy-reporting is [10, 11]. often used. The completion of 24-h recalls involves prompting re- What are the key feasibility findings? The response spondents to report all foods and drinks consumed ei- rate for the ASA24 recall was 35%. Of those ther over the past 24 h or over the previous calendar completers who were asked to perform a second day, and traditionally have been conducted by highly recall, the response rate was 59%. Few responses (4% trained interviewers, with trained coders needed to link of first recalls) were excluded from the analysis for the data to food composition databases [12]. In recent ineligibility, missing data, or inadvertent parental years, innovative tools have been developed to improve self-report. The median amount of time to complete the feasibility of collecting recalls in large-scale research, the first recall was 29.5 min (interquartile range 17.0 such as epidemiologic cohorts. One such tool is the Au- min). Further, nutrient intakes for energy, total fat, tomated Self-Administered 24-H Dietary Assessment protein, carbohydrates, fiber, sodium, total sugars, Tool (ASA24), which was developed by the US National and added sugars were similar between the ASA24 Cancer Institute based on the multiple-pass method and an interview-administered 24-h recall from the used in national surveillance [13], including the Canad- Canadian Community Health Survey. ian Community Health Survey (CCHS) [14]. The ASA24 What are the implications of the feasibility findings system is freely available for use by researchers and has for the design of the main study? The findings from been adapted specifically for use in Canada, with adapta- our study suggest that the ASA24 is an overall tions to the Canadian food supply and linkages to the feasible approach for collecting parent proxy- Canadian Nutrient File [15]. reported recall data among young children. This will The literature shows that ASA24 recalls are feasible create opportunities for its use as a dietary assess- and perform well relative to true intake and interviewer- ment tool within larger cohort studies aiming to administered recalls in samples of adults [16, 17]. Over understand the relationship between dietary factors recent years, several studies have assessed the feasibility and long-term health outcomes. of its use among older children, namely from ages 10 to 13 years [15, 18–22]. However, there has been relatively little research to inform the feasibility of ASA24 to as- Background sess dietary intake in young children. Specific consider- Measuring dietary intake in childhood is important for ations related to collecting intake data for children evaluating population health, as well as for assessing asso- include their cognitive stage, literacy and numeracy ciations between patterns of eating and health outcomes skills, and knowledge of foods and their preparation [1, 2]. Childhood eating patterns may track into later life [23–25]. Parents or other proxy reporters are often [3, 4], with implications for health and disease risk across called upon to report young children’s intake, with 10 the life course. For instance, poor eating patterns in early years suggested as the age at which children may be able life and childhood has been associated with adulthood car- to report independently [25]. One small validation feed- diovascular disease [5]. Furthermore, the 2016 Global Bur- ing study with preschoolers found that parents were able den of Diseases, Injuries, and Risk Factors Study found to report their children’s intake for a single day relatively that in Canada, dietary risk was the greatest attributable well using ASA24 [26], but the results do not provide in- factor to death, representing 17.6% of total deaths [6]. sights into the feasibility of ASA24 for use in Consequently, the accurate measurement of childhood community-based samples.
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 3 of 10 The objective of this study was thus to examine the somewhat unannounced, which is recommended to re- feasibility of use of ASA24-Canada-2016 recalls to meas- duce the likelihood that parents will alter their children’s ure parent proxy-reported dietary intake among children eating patterns due to social desirability bias [28]. If the four years of age and older. The secondary objective was recall was not completed, the research assistant sent one to compare intake estimates from ASA24 with those reminder email after 1 week, and a second reminder yielded by national surveillance data collected using after 2 weeks. interviewer-administered multiple pass 24-h recalls. Since recalls capture intake for a given day, large sur- veys often collect a second non-consecutive recall from Methods a subsample to allow modeling of distributions of usual Study design and respondents intake, which entails partitioning within- and between- A feasibility study was conducted between March 3rd, person variation and removing the within-person vari- 2018 and June 4th, 2019 to describe the use of ASA24- ation (which mostly comes from day-to-day variation) Canada-2016 among children participating in The Ap- such that inferences can be made about proportions plied Research Group for Kids (TARGet Kids!) cohort above and below recommendations, for example [29]. study. TARGet Kids! is a primary care practice-based re- Thus, approximately 25% of respondents who completed search network for children primarily in the Greater To- the first recall were asked to complete a second recall ronto Area, Ontario, Canada [27]. On an ongoing basis, about 2 weeks later, with reminder emails again sent children under six years of age are recruited into TAR- after 1 week and after 2 weeks to respondents who had Get Kids! from primary care practices (pediatrics or fam- not yet completed. ily medicine) by trained research assistants and followed When the feasibility study began in March 2018, no in- prospectively throughout childhood and adolescence at centive was provided to parents; however, due to low re- annual well-child visits. Children were excluded at en- sponse rate, an incentive was introduced on November rollment if they were < 32 weeks gestational age, had 28th, 2018 to encourage completion. Parents were in- non-English-speaking parents, had growth-restricting formed in the email asking them to complete ASA24 that health conditions such as cystic fibrosis or failure to they would receive a $5 electronic gift card (for a coffee thrive, severe developmental delay, or other chronic con- shop) for both the first and second recall. During the 16- ditions (not including asthma and high-functioning aut- month period of this feasibility study, 8 months of data ism). Over 10,500 children have been recruited into collection were conducted with no incentive, and the sub- TARGet Kids! since its inception in 2008. For this feasi- sequent 8 months were conducted with the incentive. bility study, children were included if they completed a In addition to the interface used by respondents to TARGet Kids! visit (either baseline or follow-up) be- complete the recall, ASA24 includes a researcher website tween March 2018 and June 2019 and were at least 4 from which recall data are downloaded. These data were years of age at the time of the visit. This lower age cutoff linked by study identification number to survey and an- was chosen to avoid overburdening the parents of youn- thropometric data from the most recent TARGet Kids! ger children in the TARGet Kids! cohort who are asked visit. Parent-completed survey data included child age and to complete several other age-specific questionnaires. sex, maternal ethnicity, and family income. Child height No upper age limit was applied. and weight were measured by trained research assistants, and body mass index (BMI) z scores were calculated using Data collection the recommended World Health Organization Standards Parents were sent a standardized email by a trained re- and Reference [30]. search assistant requesting online completion of ASA24- Parents provided written consent for participation in Canada-2016 for their child. The email included a link TARGet Kids!. Research ethics board approval was ob- to the ASA24 and a username and password. Parents tained from the Hospital for Sick Children, St. Michael’s completed ASA24 independently at home (or their Hospital, Toronto, Ontario and the Hamilton Research choice of location). When parents logged in to ASA24, Ethics Board, Hamilton, Ontario. they were asked to report all foods and drinks consumed within the previous 24 h from the time of completion. Data cleaning Within the email, parents were requested to “remember Each food and beverage reported by respondents was to complete it only for your child’s dietary intake (not for automatically coded within ASA24 based on a database yourself)”, since ASA24 queries respondents about their adapted from the Canadian Nutrient File [31]. ASA24 own intake and does not offer an option to change captured portion size by prompting respondents to prompts to facilitate proxy-reporting. Parents were asked choose between various images of the food items. For to complete ASA24 for either the day the email was sent each respondent, details of each individual food and bev- or the day after such that the recall was at least erage item reported as well as nutrient intake for macro-
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 4 of 10 and micronutrients for each item were downloaded from the present analyses). Data from children aged 4 to 15 the ASA24 researcher website. As a note, the ASA24 years (n = 4124) were included to match the age range of does not capture discretionary salt use given that most the TARGet Kids! respondents for whom ASA24 data dietary salt intake comes from processed foods. Guide- were available. Statistics Canada’s survey weights were ap- lines from the US National Cancer Institute were applied plied to the CCHS analyses and confidence intervals were for data cleaning [32]. Reports identified as breakoffs estimated using bootstrapping with 500 balanced repeated (i.e., the participant entered some food items but exited replications to account for the complex sampling design ASA24 before reaching the final question) were removed [14]. All statistical analyses were performed using SAS due to missing data. The data were manually scanned Studio 3.71 (SAS Institute Inc., Cary, NC, USA). for open-ended text entries that required review or du- plicate food item entries, but none were identified. Add- Results itionally, entries considered to be implausible for Figure 1 outlines the flow of respondents throughout the children, such as wine, espresso, > 1 g of alcohol, or > study. During the study period, 471 parents with chil- 50 g of caffeine were manually inspected to evaluate dren in the TARGet Kids! cohort were invited to whether intake was inadvertently recalled for the parent complete ASA24. A total of 163 (35%) parents contacted instead of the child. This inspection included an assess- by email completed a first recall for their children. Seven ment of the child’s age, as some older children in the (4%) recalls completed by these respondents were ex- sample may consume alcohol or caffeine themselves, but cluded because children were found to not meet the eli- all of the children identified with these alcohol and caf- gibility criteria (below 4 years of age; n = 2), or parents feine values were younger (< 9 years of age), and thus were suspected to have reported their own rather than these responses were excluded. the children’s intake (n = 5), leaving 156 children with data for the first recall. Four recalls included five or Statistical analysis fewer foods, but the parents indicated that the reported The examination of feasibility focused on response rates foods and beverages represented usual intake for their for each recall, number of attempts (i.e., logins to the children; the data from these recalls were therefore online system) needed before the recall was completed, retained. There were 308 non-respondents. Out of these, number of foods entered, and time to complete the re- 301 were included in the non-respondent analysis; seven call. The response rate for each recall was compared (2%) were removed for not meeting the study eligibility pre- and post-introduction of the incentive. Further, we criteria (below 4 years of age). examined whether the recall was indicative of typical A total of 78 parents who completed the first recall dietary intake based on the final ASA24 question, which were asked to complete a second recall. Forty-six (59%) asks participants if their intake that day was “much less second recalls were completed and one (2%) was ex- than usual intake”, “usual intake”, or “much more than cluded as a breakoff, leaving 45 children with data from usual intake”. Descriptive statistics were calculated as a second recall. None of the second recalls appeared in- the mean and standard deviation (SD) or the median complete based on the number of foods reported. and interquartile range (IQR) for continuous measures Table 1 describes the demographic characteristics of and as frequencies (%) for categorical measures. Descrip- the sample completing each of the recalls, as well as tive statistics were used to characterize the study popula- non-respondents. Among those completing the first re- tion, including child age, sex, and number of siblings, call, the mean age of the children was 8.7 years (SD 3.0); maternal ethnicity, family income, and BMI z score. The the age range was 4.1 to 15.3 years of age (IQR 5.7). demographic characteristics of ASA24 non-respondents Among those for whom a second recall was available, compared with those of the respondents. the mean age was 8.9 years (SD 2.9) with a range of 4.1 Descriptive statistics, including means (95% confidence to 14.0 (IQR 4.3). The distribution of children’s ages was interval (CI)), and medians (IQR), were calculated for similar between those who did and did not have ASA24 eight nutrients: energy, total fat, protein, carbohydrates, recall data available; however, compared with non- fiber, sodium, total sugars, and added sugars. These esti- respondents, parents completing ASA24 were more mates were compared with estimates from the CCHS likely to have male children, have a single child, report a 2015-Nutrition share file [33] using the 95% CIs. The higher family income, and to be of European maternal 2015 CCHS included 20,487 Canadians over 1 year of age ethnicity. Those completing ASA24 were also less likely and collected intake data by interviewer-administered 24- to have children with BMI z scores > 2 (reflective of h recalls completed using the same multiple-pass method obesity). that informed ASA24 [13, 14]. For the purposes of esti- Table 2 describes the completion characteristics asso- mating mean and median intakes, the first recall was used ciated with the first recall and the second recall. The re- (second recalls available for a subsample were not used in sponse rate for the first recall was 33.8% pre-incentive
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 5 of 10 Fig. 1 Flow chart of respondents throughout study, showing reasons for exclusion and the final sample sizes for both the first and second recalls and 36.3% post-incentive. For the second recall, the re- participants, the mean estimates were very similar with sponse rate was 51.3% pre-incentive and 66.7% post- overlapping 95% CI for energy, protein, total fat and so- incentive. The median number of days between receiving dium, and only slightly higher for carbohydrates and an invitation to complete the first recall and completing total sugars and slightly lower for fiber. Based on the it was 2.0 (IQR 7.5). The median number of days be- 95% CI, there was evidence of statistically significant dif- tween completing the first recall and receiving an invita- ferences between the first ASA24 recall and the CCHS tion for the second recall was 21.0 (IQR 13.0). For the for carbohydrates, fiber, and total sugars, but the ASA24 second recall, the median time between receiving an in- estimates were somewhat similar. vitation and completing was 7.0 days (IQR 7.0). The median number of foods reported was similar for Discussion the first and second recalls, at 18.0 (IQR 6.0) for the first This is one of the first studies to assess the feasibility of recall and 18.0 (IQR 8.0) for the second. By comparison, parent proxy-reported childhood dietary intake using the median number of foods reported on recalls col- ASA24. Overall, ASA24 appears to be a feasible ap- lected within the 2015 CCHS was 14.0 (IQR 4.8); this proach for proxy-reported dietary assessment in children difference may be partially attributed to differences in as young as 4 years of age. This finding aligns with that how multi-ingredient foods are coded in ASA24 versus of Trolle and colleagues [34], who assessed the feasibility CCHS. The median completion time for ASA24 was of parent proxy-reported in-person 24-h dietary recalls 29.5 (IQR 17.0) minutes for the first recall and 23.0 for a small group of 4–5-year-old children in Denmark (IQR 14.0) minutes for the second. For both the first and Spain. In that study, based on evaluation question- and second recalls, over 90% of respondents completed naires provided after completion of the recalls, less than ASA24 in one attempt and indicated the foods reported 15% of parents found it ‘difficult’ or ‘a little difficult’ to for that day reflected their child’s usual intake. Recalls describe and report their child’s dietary intake; most were collected from all days of the week although re- found it ‘fairly easy’ or ‘easy’ [34]. The present study’s re- spondents were more likely to complete for weekdays sponse rates, 34.6% for the first recall and 59.0% for the (89% of first recalls and 91% of second recalls) compared second, were similar to those of other studies in which with weekends (11% of first recalls and 9% of second adult members of an existing cohort were asked to recalls). complete additional dietary measures [35, 36]. Illner Table 3 shows the nutrient intake as estimated from et al. [35] conducted dietary assessments with a random the first and second ASA24 recalls. The values for all sample of five existing European cohorts, with an overall nutrients were similar between the first and second re- participation rate of 65.3% (participation rates for each calls. Further, when the TARGet Kids! first ASA24 recall cohort ranged from 37.5 to 87.5%). The response rate was compared with national estimates among CCHS for the 2015 CCHS was also similar, at 61.6% [14].
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 6 of 10 Table 1 Demographic characteristics of samples completing first (N = 156) and second (N = 45) recalls, and of non-respondents (N = 301) Characteristic Non-respondents First recall (N = 156) Second recall (N = 45) (N = 301) Age (years)a, mean (SD) 8.8 (2.9) 8.7 (3.0) 8.9 (2.9) 4–6 (n, %) 78 (25.9) 45 (28.8) 11 (24.4) 7–10 (n, %) 138 (45.8) 65 (41.7) 20 (44.4) 11–15 (n, %) 85 (28.2) 46 (29.5) 14 (31.1) Sexa, n (%) Female 156 (51.8) 67 (43.0) 20 (44.4) Male 145 (48.2) 89 (57.1) 25 (55.6) Siblingsa, n (%) 0 30 (16.0) 21 (19.4) 7 (24.1) 1 115 (61.2) 64 (59.3) 16 (55.2) 2 or more 43 (22.9) 23 (21.3) 6 (20.7) Missingb 113 48 16 Family income, n (%) Less than $30,000 3 (1.6) 2 (1.9) 1 (3.6) $30,000 to $79,999 27 (14.8) 8 (7.5) 3 (10.7) $80,000 to $149,999 49 (26.8) 38 (35.5) 10 (35.7) $150,000 or more 104 (56.8) 59 (55.1) 14 (50.0) Missingb 118 49 17 BMI z scorea, n (%) ≤ 1 (normal or underweight) 226 (75.6) 126 (81.8) 34 (75.6) > 1 and ≤ 2 (overweight) 46 (15.4) 20 (13.0) 9 (20.0) > 2 (obesity) 27 (9.0) 8 (5.2) 2 (4.4) Missing 2 2 0 Maternal ethnicity, n (%) European 182 (67.9) 104 (73.2) 31 (77.5) Asian 46 (17.2) 20 (14.1) 6 (15.0) Otherc 40 (14.9) 18 (12.7) 3 (7.5) Missing 33 14 5 a Of the child (not the parent) b Data on siblings and income were collected from the yearly age-specific questionnaires, which were not yet available for all participants. c Includes Arab, African, Latin American, mixed ethnicity, and other Measuring the frequency of proxy under- and over- Our study included children from 4 to 15 years of age reporting of dietary intake was not possible in the (70% were ≤ 10 years of age), and parents were asked to current study. However, in a standardized daycare set- complete the recall for their child. The literature sug- ting, parent proxy-reporting of the foods and beverages gests that proxy-reporting should be used for children consumed by 2–5-year olds using ASA24 was found to below 10 years of age [25]. A validation study specific to have 80% close or exact matches when compared with the ASA24 suggested that children between the ages of true dietary intake as measured through direct observa- 9 and 11 years may not be able to successfully complete tion [26]. B rnhorst et al. [37] assessed the prevalence the recall alone, and that more research is needed to de- of misreporting for parent proxy-report of dietary intake termine the age at which parental assistance will no lon- using a European online 24-h recall for children ages 2– ger be required [20]. For these older children, it may be 9 years. They found that the prevalence of under- and more appropriate to encourage parents to complete the over-reporting of intake was 8.0% and 3.4%, respectively, ASA24 with their children (i.e., proxy-assisted respond- when compared with Goldberg cutoff values for age-, ing). The CCHS administers proxy-reporting of dietary sex-, and body size-specific basal metabolic rate [37]. intake for children 1–5 years of age, with parent-assisted
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 7 of 10 Table 2 Completion characteristics for the first (N = 156) and second (N = 45) recalls Characteristics of completion First recall (N = 156) Second recall (N = 45) Response rate overall, n (%) 163/471 (34.6) 46/78 (59.0) Response rate pre-incentive, n (%) 105/311 (33.8) 20/39 (51.3) Response rate post-incentive, n (%) 58/160 (36.3) 26/39 (66.7) Number of attempts to complete, n (%) 1 144 (92.3) 45 (100) 2 12 (7.7) 0 (0) Recall indicative of normal intake, n (%) Much more than usual intake 4 (2.6) 0 (0) Usual intake 144 (92.3) 43 (95.6) Much less than usual intake 8 (5.1) 2 (4.4) Time between receiving email invitation and completing recall (days), median (IQR) 2.0 (7.5) 7.0 (7.0) Total recall session duration (min), median (IQR) 29.5 (17.0) 23.0 (14.0) Number of foods reported, median (IQR) 18.0 (6.0) 18.0 (8.0) recalls for 6–12, and non-proxy (child completion) for completion of the ASA24 among low-income adults, as those aged 12 years and up [33]. Although our study well as adults from certain ethnic groups [38, 39]; how- did not collect data on whether the child was present ever, this has not been found in all studies [40]. while their parent completed the recall, this informa- Although we observed differences between the charac- tion may have clarified the benefits of proxy-reporting teristics of respondents and non-respondents, estimated compared with parent-assisted reporting among chil- nutrient intakes were very similar to national estimates dren of varying ages. from the CCHS, which is a large, nationally representa- Systematic differences were observed between respon- tive survey. Overall, efforts are needed to increase dents and non-respondents on key characteristics, in- response rates and support completion across demo- cluding child sex, number of siblings, family income graphic subgroups to enhance the generalizability of esti- level, maternal ethnicity, and BMI z score. These differ- mates yielded by cohorts, such as TARGet Kids!. For ences create the potential for selection bias, which can example, a systematic probability-based sampling tech- reduce generalizability and, in studies evaluating associa- nique could be used to improve generalizability. To in- tions between diet and health outcomes, can lead to crease response rate, the use of multiple methods to biased measures of association. This finding is consistent contact parents, including email, text messaging, and with previous research that has identified lack of phone calls, may be beneficial. Table 3 Mean intake from the first (N = 156) and second (N = 45) recalls, compared with the CCHSa (N = 4124) Nutrient Unit First recall Second recall CCHS first recall (N = 156) (N = 45) (N = 4124) Mean Median (IQR) Mean Median (IQR) Mean Median (IQR) (95%CI) (95%CI) (95%CI) Energy (kcal/day) 1742.6 1704.0 (840.2) 1788.1 1740.3 (650.9) 1850.7 1728.0 (793.3) (1637.3–1847.9) (1635.8–1940.4) (1742.0–1959.4) Protein (g/day) 70.5 (65.6–75.4) 65.0 (35.7) 75.1 (67.0–83.2) 74.3 (46.4) 70.6 (67.6–73.7) 64.0 (37.0) Total fat (g/day) 65.6 (60.2–71.0) 60.1 (39.4) 68.3 (59.6–77.0) 61.9 (39.1) 64.9 (58.9–70.9) 58.6 (36.7) Carbohydrates (g/day) 225.4 213.7 (98.5) 224.9 220.5 (100.2) 251.7 235.1 (111.5) (211.6–239.2) (204.8–245.0) (240.1–263.3) Fiber (g/day) 18.3 (16.9–19.7) 17.1 (10.1) 17.4 (15.6–19.2) 16.5 (6.8) 15.6 (15.2–16.0) 14.1 (8.7) Sodium (mg/day) 2654.1 2498.7 (1472.0) 2709.8 2493.2 (1244.0) 2594.9 2357.9 (1393.5) (2465.3–2842.9) (2434.7–2984.9) (2400.9–2788.8) Total sugars (g/day) 96.2 93.5 (54.6) 95.3 95.0 (53.4) 111.5 101.5 (63.0) (89.0–103.4) (85.3–105.3) (103.8–119.1) Added sugars (g/day) 38.9 (34.2–43.6) 35.0 (31.9) 37.8 (30.3–45.3) 30.7 (28.3) Not reported Not reported a Twenty-four-hour recall data collected within the 2015 CCHS (Canadian Community Health Survey) for children aged 4 to 15 years
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 8 of 10 This study offers valuable insights with regards to com- TARGet Kids!: The Applied Research Group for Kids; BMI: Body mass index; pletion of new assessments by current members of co- SD: Standard deviation; CI: Confidence interval; IQR: Interquartile range horts. Demographic data were available on non- respondents through the TARGet Kids! cohort, enabling Supplementary Information comparison characteristics of respondents versus non- The online version contains supplementary material available at https://doi. org/10.1186/s40814-021-00864-6. respondents; many studies are not able to describe non- respondents. We were also able to evaluate changes in the Additional file 1. CONSORT 2010 checklist of information to include response rate following introduction of an incentive due when reporting a pilot or feasibility trial*. to a change in study protocol mid-study. Notably, survey Additional file 2. STROBE Statement—Checklist of items that should be participation improved minimally with the introduction of included in reports of cohort studies. the incentive. This finding is in line with existing literature on survey response rates, which suggests that prepaid and Acknowledgments cash-type incentives are more effective than promised and We thank all of the participating families for their time and involvement in TARGet Kids! and are grateful to all practitioners who are currently involved gift card-type incentives [41, 42]. in the TARGet Kids! practice-based research network. TARGet Kids! collabora- However, several considerations are relevant to the in- tors—co-leads: Catherine S. Birken and Jonathon L. Maguire; Advisory Com- terpretation of our findings. ASA24 was not designed for mittee: Ronald Cohn, Eddy Lau, Andreas Laupacis, Patricia C. Parkin, Michael Salter, Peter Szatmari, and Shannon Weir; Science Review and Management proxy-reporting and could not be customized to ask spe- Committees: Laura N. Anderson, Cornelia M. Borkhoff, Charles Keown- cifically about the foods and drinks that “your child had” Stoneman, Christine Kowal, and Dalah Mason; Site Investigators: Murtala rather than the foods and drinks that “you had”. Five re- Abdurrahman, Kelly Anderson, Gordon Arbess, Jillian Baker, Tony Barozzino, Sylvie Bergeron, Dimple Bhagat, Gary Bloch, Joey Bonifacio, Ashna Bowry, calls were excluded as it appeared, due to high amounts Caroline Calpin, Douglas Campbell, Sohail Cheema, Elaine Cheng, Brian Chi- of caffeine or alcohol, the parents likely reported their samore, Evelyn Constantin, Karoon Danayan, Paul Das, Mary Beth Derocher, own intake rather than their child’s. We cannot rule out Anh Do, Kathleen Doukas, Anne Egger, Allison Farber, Amy Freedman, Sloane Freeman, Sharon Gazeley, Charlie Guiang, Dan Ha, Curtis Handford, Laura the possibility that other parents may have inadvertently Hanson, Leah Harrington, Sheila Jacobson, Lukasz Jagiello, Gwen Jansz, Paul reported their own intake as well. A version of ASA24 Kadar, Florence Kim, Tara Kiran, Holly Knowles, Bruce Kwok, Sheila Lakhoo, that includes the option for proxy-reporting would Margarita Lam-Antoniades, Eddy Lau, Denis Leduc, Fok-Han Leung, Alan Li, Patricia Li, Jessica Malach, Roy Male, Vashti Mascoll, Aleks Meret, Elise Mok, broaden its applicability to studies with young children Rosemary Moodie, Maya Nader, Katherine Nash, Sharon Naymark, James and other populations that may benefit from proxy- Owen, Michael Peer, Kifi Pena, Marty Perlmutar, Navindra Persaud, Andrew reporting. This study did not include any assessment of Pinto, Michelle Porepa, Vikky Qi, Nasreen Ramji, Noor Ramji, Danyaal Raza, Alana Rosenthal, Katherine Rouleau, Caroline Ruderman, Janet Saunderson, the usability of the ASA24 tool itself, hindering our un- Vanna Schiralli, Michael Sgro, Hafiz Shuja, Susan Shepherd, Barbara Smiltnieks, derstanding of parent’s experiences with completing Cinntha Srikanthan, Carolyn Taylor, Stephen Treherne, Suzanne Turner, Fatima ASA24 on behalf of their child and potential barriers to Uddin, Meta van den Heuvel, Joanne Vaughan, Thea Weisdorf, Sheila Wijaya- singhe, Peter Wong, John Yaremko, Ethel Ying, Elizabeth Young, and Michael completion as well as strategies to overcome these bar- Zajdman; Research Team: Farnaz Bazeghi, Vincent Bouchard, Marivic Bustos, riers. Other studies have assessed the usability of ASA24 Charmaine Camacho, Dharma Dalwadi, Christine Koroshegyi, Tarandeep with measures such as the System Usability Scale [15]. Malhi, Sharon Thadani, Julia Thompson, and Laurie Thompson; Project Team: Mary Aglipay, Imaan Bayoumi, Sarah Carsley, Katherine Cost, Karen Eny, Ther- Lastly, our study represents an urban population of chil- esa Kim, Laura Kinlin, Jessica Omand, Shelley Vanderhout, and Leigh Vander- dren from Canada and may not be generalizeable to loo; Applied Health Research Centre: Christopher Allen, Bryan Boodhoo, Olivia other populations. Chan, David W.H. Dai, Judith Hall, Peter Juni, Gerald Lebovic, Karen Pope, and Kevin Thorpe; Mount Sinai Services Laboratory: Rita Kandel, Michelle Rodri- gues, and Hilde Vandenberghe. Conclusions Authors’ contributions Many studies involving young children rely on dietary SIK, BS, JLM, CSB, and LNA conceptualized the study. IS, CKS, JO, and SV screening questionnaires or short FFQs, which tend to contributed to the data analysis and interpretation. IS and LNA wrote the have higher levels of systematic error, or bias, compared first draft of the manuscript, and all authors contributed to revising the manuscript, and the authors read and approved the final manuscript. with 24-h recalls [10, 11]. Overall, this study suggests ASA24 is a feasible approach for collecting recall data Funding among young children, opening up possibilities for im- Funding for this research was obtained from the Canadian Institutes of proved assessment of dietary intake in cohorts aiming to Health Research (CIHR). The funding body had no role in the study design, collection, analysis, interpretation, or writing. understand how dietary intake and other factors influ- ence long-term health. However, further research should Availability of data and materials consider strategies to overcome barriers to completion The data that support the findings of this study are available from TARGet among all cohort members. Kids!, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available upon request with approval of the TARGet Kids! Abbreviations Scientific Committee and institutional research ethics boards by contacting FFQ: Food Frequency Questionnaire; ASA24: Automated Self-Administered http://www.targetkids.ca/contact-us/. The CCHS data that were used in this 24-H Dietary Assessment Tool; CCHS: Canadian Community Health Survey; study are available to authorized users through Statistics Canada.
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 9 of 10 Declarations 9. Satija A, Yu E, Willett WC, Hu FB. Understanding nutritional epidemiology and Its role in policy. Adv Nutr. 2015;6(1):5–18. https://doi.org/10.3945/a Ethics approval and consent to participate n.114.007492. Parents provided written consent for participation in TARGet Kids!. Research 10. Freedman LS, Commins JM, Moler JE, Arab L, Baer DJ, Kipnis V, et al. Pooled ethics board approval was obtained from the Hospital for Sick Children, St. Results From 5 validation studies of dietary self-report Instruments using Michael’s Hospital, Toronto Ontario and the Hamilton Research Ethics Board, recovery biomarkers for energy and protein intake. Am J Epidemiol. 2014; Hamilton, Ontario. 180(2):172–88. https://doi.org/10.1093/aje/kwu116. 11. Freedman LS, Commins JM, Moler JE, Willett W, Tinker LF, Subar AF, et al. Pooled results from 5 validation studies of dietary self-r Instruments Using Consent for publication Recovery Biomarkers for Potassium and Sodium Intake. Am J Epidemiol. Not applicable 2015;181(7):473–87. https://doi.org/10.1093/aje/kwu325. 12. Thompson FE, Byers T. Dietary Assessment Resource Manual. J Nutr. 1994; 124(suppl_11):2245s–317s. Competing interests 13. Subar AF, Kirkpatrick SI, Mittl B, Zimmerman TP, Thompson FE, Bingley C, JLM received an unrestricted research grant for a completed investigator- et al. The Automated Self-Administered 24-Hour Dietary Recall (ASA24): a initiated study from the Dairy Farmers of Canada (2011–2012), and Ddrops resource for researchers, clinicians, and educators from the National Cancer provided non-financial support (vitamin D supplements) for an investigator- Institute. J Acad Nutr Diet. 2012;112(8):1134–7. https://doi.org/10.1016/j.ja initiated study on vitamin D and respiratory tract infections (2011–2015). IS, nd.2012.04.016. SIK, BS, CDGKS, JO, SV, and LNA declare that they have no competing 14. Health Canada: Reference Guide to Understanding and Using the Data: interests. 2015 Canadian Community Health Survey—Nutrition. https://deslibris.ca/ ID/10093153 (2017). Author details 1 15. Kirkpatrick SI, Gilsing AM, Hobin E, Solbak NM, Wallace A, Haines J, et al. Department of Health Research Methods, Evidence, and Impact, McMaster Lessons from Studies to Evaluate an Online 24-Hour Recall for Use with University, Hamilton, Ontario, Canada. 2School of Public Health and Health Children and Adults in Canada. Nutrients. 2017;9(2). Systems, University of Waterloo, Waterloo, Ontario, Canada. 3Department of 16. Kirkpatrick SI, Subar AF, Douglass D, Zimmerman TP, Thompson FE, Kahle Health Promotion, Chronic Disease and Injury Prevention, Public Health LL, et al. Performance of the Automated Self-Administered 24-hour Recall Ontario, Toronto, Ontario, Canada. 4Division of Epidemiology, Dalla Lana relative to a measure of true intakes and to an interviewer-administered 24- School of Public Health, University of Toronto, Toronto, Ontario, Canada. 5 h recall123. Am J Clin Nutr. 2014;100(1):233–40. https://doi.org/10.3945/a Applied Health Research Centre of the Li Ka Shing Knowledge Institute of jcn.114.083238. St. Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada. 6 17. Thompson FE, Dixit-Joshi S, Potischman N, Dodd KW, Kirkpatrick SI, Kushi Division of Biostatistics, Dalla Lana School of Public Health, University of LH, et al. Comparison of interviewer-administered and automated self- Toronto, Toronto, Ontario, Canada. 7Division of Child Health Evaluative administered 24-hour dietary recalls in 3 diverse integrated health systems. Sciences (CHES), Sick Kids Research Institute, Toronto, Ontario, Canada. 8 Am J Epidemiol. 2015;181(12):970–8. https://doi.org/10.1093/aje/kwu467. Department of Nutritional Sciences, Faculty of Medicine, University of 18. Raffoul A, Hobin EP, Sacco JE, Lee KM, Haines J, Robson PJ, et al. School-age Toronto, Toronto, Ontario, Canada. 9Department of Pediatrics, St. Michael’s children can recall some foods and beverages consumed the Prior Day Hospital, Toronto, Ontario, Canada. Using the Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24) without Assistance. J Nutr. 2019;149(6):1019–26. https://doi.org/10.1 Received: 15 March 2021 Accepted: 3 June 2021 093/jn/nxz013. 19. Hughes AR, Summer SS, Ollberding NJ, Benken LA, Kalkwarf HJ. Comparison of an interviewer-administered with an automated self-administered 24 h References (ASA24) dietary recall in adolescents. Public Health Nutr. 2017;20(17):3060–7. 1. World Cancer Research Fund/American Institute for Cancer Research. https://doi.org/10.1017/S1368980017002269. Recommendations and public health and policy implications. Continuous 20. Diep CS, Hingle M, Chen T-A, Dadabhoy HR, Beltran A, Baranowski J, et al. A Update Project Expert Report. 2018. dietandcancerreport.org. validation study of the Automated Self-Administered 24-Hour Dietary Recall 2. Afshin A, Sur PJ, Fay KA, Cornaby L, Ferrara G, Salama JS, et al. Health effects for Children (ASA24-Kids) among 9 to 11-year-old youth. J Acad Nutr Diet. of dietary risks in 195 countries, 1990–2017: a systematic analysis for the 2015;115(10):1591–8. https://doi.org/10.1016/j.jand.2015.02.021. Global Burden of Disease Study 2017. The Lancet. 2019;393(10184):1958–72. 21. Krehbiel CF, DuPaul GJ, Hoffman JA. A validation study of the Automated https://doi.org/10.1016/S0140-6736(19)30041-8. Self-Administered 24-Hour Dietary Recall for Children, 2014 Version, at 3. Mikkilä V, Räsänen L, Raitakari OT, Pietinen P, Viikari J. Consistent dietary School Lunch. J Acad Nutr Diet. 2017;117(5):715–24. https://doi.org/10.1016/ patterns identified from childhood to adulthood: the cardiovascular risk in j.jand.2016.10.028. Young Finns Study. Br J Nutr. 2005;93(6):923–31. https://doi.org/10.1079/ 22. Baranowski T, Islam N, Baranowski J, Martin S, Beltran A, Dadabhoy H, et al. BJN20051418. Comparison of a web-based versus traditional diet recall among 4. Movassagh EZ, Baxter-Jones ADG, Kontulainen S, Whiting SJ, Vatanparast H. children. J Acad Nutr Diet. 2012;112(4):527–32. https://doi.org/10.1016/j. Tracking dietary patterns over 20 years from childhood through jada.2011.10.002. adolescence into young adulthood: the Saskatchewan pediatric bone 23. Smith AF, Baxter SD, Hitchcock DB, Finney CJ, Royer JA, Guinn CH. Cognitive mineral accrual study. Nutrients. 2017;9(9). ability, social desirability, body mass index, and socioeconomic status as 5. Kaikkonen JE, Mikkilä V, Magnussen CG, Juonala M, Viikari JSA, Raitakari OT. correlated of fourth-grade children’s dietary-reporting accuracy. Eur J Clin Does childhood nutrition influence adult cardiovascular disease Nutr. 2016;70(9):1028–33. https://doi.org/10.1038/ejcn.2016.43. risk?—Insights from the Young Finns Study. Ann Med. 2013;45(2):120–8. 24. Baxter SD, Smith AF, Litaker MS, Guinn CH, Shaffer NM, Baglio ML, et al. https://doi.org/10.3109/07853890.2012.671537. Recency affects reporting accuracy of children’s dietary recalls. Ann Epidemiol. 6. Alam S, Lang JJ, Drucker AM, Gotay C, Kozloff N, Mate K, et al. Assessment 2004;14(6):385–90. https://doi.org/10.1016/j.annepidem.2003.07.003. of the burden of diseases and injuries attributable to risk factors in Canada 25. Livingstone MBE, Robson PJ, Wallace JMW. Issues in dietary intake from 1990 to 2016: an analysis of the Global Burden of Disease Study. CMAJ assessment of children and adolescents. Br J Nutr. 2004;92(S2):S213–22. Open. 2019;7(1):E140–8. https://doi.org/10.9778/cmajo.20180137. https://doi.org/10.1079/BJN20041169. 7. Thompson FE, Kirkpatrick SI, Krebs-Smith SM, Reedy J, Schap TE, Subar AF, 26. Wallace A, Kirkpatrick SI, Darlington G, Haines J. Accuracy of parental et al. The National Cancer Institute’s Dietary Assessment Primer: a resource reporting of preschoolers’ dietary intake using an online self-administered for diet research. J Acad Nutr Diet. 2015;115(12):1986–95. https://doi.org/1 24-h Recall. Nutrients. 2018;10(8):987. https://doi.org/10.3390/nu10080987. 0.1016/j.jand.2015.08.016. 27. Carsley S, Borkhoff CM, Maguire JL, Birken CS, Khovratovich M, McCrindle B, 8. Subar AF, Freedman LS, Tooze JA, Kirkpatrick SI, Boushey C, Neuhouser ML, Macarthur C, Parkin PC, TARGet Kids! Collaboration. Cohort Profile: The et al. Addressing current criticism regarding the value of self-report Dietary Applied Research Group for Kids (TARGet Kids!). Int J Epidemiol. 2015;44(3): Data12. J Nutr. 2015;145(12):2639–45. https://doi.org/10.3945/jn.115.219634. 776–788, DOI: https://doi.org/10.1093/ije/dyu123.
Sharpe et al. Pilot and Feasibility Studies (2021) 7:123 Page 10 of 10 28. Hebert JR, Clemow L, Pbert L, Ockenet IS, Ockene JK. Social desirability bias in dietary self-report may cmpromise the validity of dietary intake measures. Int J Epidemiol. 1995;24(2):389–98. https://doi.org/10.1093/ije/24.2.389. 29. Dodd KW, Guenther PM, Freedman LS, Subar AF, Kipnis V, Midthune D, et al. Statistical methods for estimating usual intake of nutrients and foods: a review of the theory. J Am Diet Assoc. 2006;106(10):1640–50. https://doi. org/10.1016/j.jada.2006.07.011. 30. Canadian Paediatric Society and Dieticians of Canada. Promoting optimal monitoring of child growth in Canada: using the new World Health Organization growth charts—Executive Summary. Paediatr Child Health. 2010;15(2):77–79. 31. Health Canada: Canadian Nutrient File Search Engine Online. https://food- nutrition.canada.ca/cnf-fce/index-eng.jsp (2018). 32. National Cancer Institute: Reviewing and Cleaning ASA24® Data. https://epi. grants.cancer.gov/asa24/resources/asa24-data-cleaning.pdf (2018). 33. Statistics Canada. CCHS 2015: Nutritional Component, General health and summary data for 24-hour dietary recall and nutritional supplements data set. ODESI. 2019. Identification number: cchs-82M0024-E-2015-nu-hs. 34. Trolle E, Amiano P, Ege M, Bower E, Lioret S, Brants H, et al. Feasibility of repeated 24-h dietary recalls combined with a food-recording booklet, using EPIC-Soft, among preschoolers. Eur J Clin Nutr. 2011;65(1):S84–6. https://doi.org/10.1038/ejcn.2011.91. 35. Illner A-K, Harttig U, Tognon G, Palli D, Salvini S, Bower E, et al. Feasibility of innovative dietary assessment in epidemiological studies using the approach of combining different assessment instruments. Public Health Nutr. 2011;14(6):1055–63. https://doi.org/10.1017/S1368980010003587. 36. Galante J, Adamska L, Young A, Young H, Littlejohns TJ, Gallacher J, et al. The acceptability of repeat Internet-based hybrid diet assessment of previous 24-h dietary intake: administration of the Oxford WebQ in UK Biobank. Br J Nutr. 2016;115(4):681–6. https://doi.org/10.1017/S000711451 5004821. 37. Börnhorst C, Huybrechts I, Ahrens W, Eiben G, Michels N, Pala V, et al. Prevalence and determinants of misreporting among European children in proxy-reported 24 h dietary recalls. Br J Nutr. 2013;109(7):1257–65. https:// doi.org/10.1017/S0007114512003194. 38. Kupis J, Johnson S, Hallihan G, Olstad DL. Assessing the usability of the Automated Self-Administered Dietary Assessment Tool (ASA24) among low- income adults. Nutrients. 2019;11(1). 39. Ettienne-Gittens R, Boushey CJ, Au D, Murphy SP, Lim U, Wilkens L. Evaluating the feasibility of utilizing the Automated Self-Administered 24- hour (ASA24) dietary recall in a sample of multiethnic older adults. Procedia Food Sci. 2013;2:134–44. https://doi.org/10.1016/j.profoo.2013.04.021. 40. Kirkpatrick SI, Guenther PM, Douglass D, Zimmerman T, Kahle LL, Atoloye A, et al. The provision of assistance does not substantially impact the accuracy of 24-hour dietary recalls completed using the Automated Self- Administered 24-H Dietary Assessment Tool among women with low incomes. J Nutr. 2019;149(1):114–22. https://doi.org/10.1093/jn/nxy207. 41. Smith MG, Witte M, Rocha S, Basner M. Effectiveness of incentives and follow-up on increasing survey response rates and participation in field studies. BMC Med Res Methodol. 2019;19(1):1–13. 42. Mercer A, Caporaso A, Cantor D, Townsend R. How much gets you how much? Monetary incentives and response rates in household surveys. Public Opinion Quarterly. 2015;79(1):105–29. https://doi.org/10.1093/poq/nfu059. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
You can also read